01001000100001001000100000110000001000001100

Robust Motor Controller Design Implemented with TMS320F240 DSP

Dr. Yongpeng Zhang Dr. Cajetan M. AkujuobiDr. Matthew SadikuDr. Tokunbo OgunfunmiCECSTR Prairie View A&M University

2

Outline

Introduction

Internal Model Control

Disturbance Rejection

Summary

3

Advantages of Direct Motor Drive

Simpler in mechanical structure, more reliable, of high performance.

4

Challenge and Solution

Challenges: without mechanical transmissionsparameter uncertaintiesdisturbance variations

DSP Solutions: digital signal processors (DSP) switch power devices software (algorithm)

5

MATLAB/SIMULINK to Programs TI TMS320F240

MATLAB/SIMULINK DSP programming (C / assembly)

6

Resulting Mixed-Signal System

Hybrid system, or sampled-data system

analog system + digital controller

yuprocessD-AA-D algorithm

clock

7

DSP-based Motor Drive System

8

Internal Model Control (IMC)

)(tr )(ty

controller

process model

)(td

q p~process

p

More stable, on-line tuning, anti-windup, etcTransformation is needed: we can not rely on model uncertainty to construct the closed-loop structure.

9

Practical IMC model

1B 1C∫

1K

1x

1A

yd

1y

)(ˆ1 sG

yΔ

y

motor

r

PID)(sGc

1u

Input SaturationModeling error

10

Physical Implementation

dSPACE DSPpower converterPMDC motor, 200W, 5A, 42V PMSM motor, 250W, 5A, 42V

11

Robust Performance

Command trackingInput saturation

12

Disturbance Rejection

i ω

LT

TKBJs +

1

disturbanceobserver

currentcompensator

LT̂

refi emT

2DOF (2-degree-of-freedom) Advantage:

satisfy the conflicted requirements of command tracking and disturbance rejection

Disadvantages:Fast depression leads to too much variation in control signal

13

Patented Novel Method

Due to load disturbance, observation error is inevitable.

)(tr )(ty

)(td

[ ]0,,, 111 CBA

observer1K

)(1 ty

=)(ˆ1 tx

motor

)()(1 tetx +

)(tu

cascadedcontroller

cG

)()()()( 11 tdJteCJAte oo +−=&

14

Feedforward Compensation

The observation error state feedback constructs a feedforward compensation for the load disturbance.The “residual disturbance” after compensation can be predicted as

r y

d

1y

eqd)(sGP)(sGc

)(ˆ sGo)(sGP

)()](ˆ)([)( sdsGsGIsd oPre ⋅−=

15

Simulated Results

Constant load disturbanceSinusoidal load disturbance

16

Experiment Results

1/8 of the simulated load torque applied

17

Patent and Publications

Motor Control Method and Apparatus with Multi-Objective Observer for Disturbance Rejection, (Application No. 60666106).

1. W. Ali, Yongpeng Zhang, C.M. Akujuobi, C.L. Tolliver, L.S. Shieh, “DSP-based PID Controller Design for the PMDC Motor”, International Journal of Modeling and Simulation, Vol. 22, No. 3, 2006, (in press).

2. Yongpeng Zhang, C.M. Akujuobi, W. Ali, L.S. Shieh, C.L. Tolliver, “Disturbance Resistance Speed Controller Design for PMSM”, IEEE Trans. On Industrial Electronics, (accepted).

3. Yongpeng Zhang, C.M. Akujuobi, L.S. Shieh and X. Gu, “Observer-based Load Disturbance Compensation for Motor Drive with DSP Implementation”, 31th Annual Conference of the IEEE Industrial Electronics Society (IECON’05), Raleigh, North Carolina, Nov 6-10, 2005.

18

Robust Motor Controller Design Implemented with TMS320F240 DSP

Dr. Yongpeng Zhang, Dr. Cajetan M. Akujuobi, Dr. Matthew Sadiku, Dr. Tokunbo Ogunfunmi

CECSTR

Prairie View A&M University